基于MRI的前庭系统分割与统计形态学分析关键技术研究

项目名称： 基于MRI的前庭系统分割与统计形态学分析关键技术研究

项目编号： No.81201157

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 影像医学与生物医学工程

项目作者： 王德峰

作者单位： 香港中文大学深圳研究院

项目金额： 24万元

中文摘要： 前庭系统是人类大脑中的平衡感知器官，许多生理或病理性刺激会引起其内部结构的解剖学变化。由于前庭系统结构复杂，其在形态方面的研究目前仅限于简单的测量与显示，很少有研究能够深入地对其形态展开分析。在青少年特发性脊柱侧弯(AIS)研究中，我们已经对前庭系统中半规管的解剖形态学异常进行了初步分析，但分析方法和精度仍需要极大的改善。因此本项目将致力于探索先进的计算技术实现前庭系统的精确自动分割与定量统计形态学分析，并在此基础上研究其解剖形态学差异与某些疾病(AIS与听力障碍)的相关性。研究以T2加权的磁共振影像为研究对象，在前庭系统统计模型基础上，逐步实现前庭系统的精确分割、高级建模、几何描述、形状分析及统计形态学测量等。该研究将会揭示前庭系统解剖形态学变化或差异与相关疾病之间的关联程度，对相关疾病的预测提供潜在的客观指导，对于内耳病理分析及内耳手术规划等也有极大的帮助。

中文关键词： 前庭系统；青少年特发性脊柱侧弯；图像分割；图像配准；

英文摘要： Vestibular system is the major balance perception organ in the human brain. Many physiological and pathological stimulation can cause anatomical changes of its internal structure. Due to the complexity in their shapes and the lack of specific image computing techniques in the past, little research has been carried out to objectively investigate the morphoanatomy of these complex structures. Our previous studies have provided pilot understanding of the relevance of morphoanatomy changes between adolescent idiopathic scoliosis patients and normal control. But further investigation of the analysis methods and accuracy is needed. This research project aims at developing a novel method to automatically extract vestibular system from magnetic resonance images and quantitatively analyze the statistical morphology. Also the correlation between the morphoanatomy differences of vestibular system and certain diseases will be deeply investigated. The research will start with the development of a statistical shape model of the vestibular system. Then base on a large number of magnetic resonance images related to certain diseases, our study will develop and conduct automatic segmentation, advanced modeling, accurate shape analysis and statistical morphometry.It will reveal the morphoanatomy of the vestibular system and the co

英文关键词： Vestibular System；AIS；Image Segmentation；Image Registration；